The Paradox of the Plankton G. E. Hutchinson the American

Home , Paradox of the plankton, Species distribution

The Paradox of the Plankton

G. E. Hutchinson

The American Naturalist, Vol. 95, No. 882. (May - Jun., 1961), pp. 137-145.

Stable URL: http://links.jstor.org/sici?sici=0003-0147%28196105%2F06%2995%3A882%3C137%3ATPOTP%3E2.0.CO%3B2-6

The American Naturalist is currently published by The University of Chicago Press.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/ucpress.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Mon Jul 2 14:28:26 2007 Vol. XCV, No. 882 The American Naturalist May-June, 1961

THEPARADOXOF THEPLANKTON*

G. E. HUTCHINSON Osborn Zoological Laboratory, New Haven, Connecticut

The problem that I wish to discuss in the present contribution is raised by the very paradoxical situation of the plankton, particularly the phytoplankton, of relatively large bodies of water. We know from laboratory experiments conducted by many workers over a long period of time (summary in Provasoli and Pintner, 1960) that most mem- bers of the phytoplankton are phototrophs, able to reproduce and build up populations in inorganic media containing a source of CO,, inorganic nitro- gen, sulphur, and phosphorus compounds and a considerable number of other elements (Na, K, Mg, Ca, Si, Fe, Mn, B, Cl, Cu, Zn, Mo, Co and V) most of which are required in small concentrations and not all of which are known to be required by all groups. In addition, a number of species are known which require one or more vitamins, namely thiamin, the cobalamines (B,, or re- lated compounds), or biotin. The problem that is presented by the phytoplankton is essentially how it is possible for a number of species to coexist in a relatively isotropic or unstructured environment all competing for the same sorts of materials. The problem is particularly acute because there is adequate evidence from en- richment experiments that natural waters, at least in the summer, present an environment of striking nutrient deficiency, so that competition is likely to be extremely severe. According to the principle of competitive exclusion (Hardin, 1960) known by many names and developed over a long period of time by many investi- gators (see Rand, 1752; Udvardy, 1757; and Hardin, 1960, for historic re- views), we should expect that one species alone would outcompete all the others so that in a final equilibrium situation the assemblage would reduce to a population of a single species. The principle of competitive exclusion has recently been under attack from a number of quarters. Since the principle can be deduced mathematically from a relatively simple series of postulates, which with the ordinary postulates of mathematics can be regarded as forming an axiom system, it follows that if the objections to the principle in any cases are valid, some or all the b'iological axioms introduced are in these cases incorrect. Most objections to the principle appear to imply the belief that equilibrium under a given set of environmental conditions is never in practice obtained. Since the deduction of the principle implies an equilibrium system, if such sys-

'Contribution to a symposium on Modern Aspects of Population Biology. Pre- sented at the meeting of the American Society of Naturalists, cosponsored by the American Society of Zoologists, Ecological Society of America and the Society for the Study of Evolution. American Association for the Advancement of Science, New York, N. Y., December 27, 1960. 138 THE AMERICAN NATURALIST tems are rarely if ever approached, the principle though analytically true, is at first sight of little empirical interest. The mathematical procedure for demonstrating the truth of the principle involves, in the elementary theory, abstraction from time. It does, however, provide in any given case a series of possible integral paths that the populations can follow, one relative to the other, and also paths that they cannot follow under a defined set of conditions. If the conditions change the integral paths change. Mere failure to obtain equilibrium owing to external vari- ation in the environment does not mean that the kinds of competition de- scribed mathematically in the theory of competitive exclusion are not oc- curing continuously in nature. Twenty years ago in a Naturalists' Symposium, I put (Hutchinson, 1941) forward the idea that the diversity of the phytoplankton was explicable pri- marily by a permanent failure to achieve equilibrium as the relevant external factors changed. I later pointed out that equilibrium would never be expected in nature whenever organisms had reproductive rates of such a kind that under constant conditions virtually complete competitive replacement of one species by another occurred in a time (t,), of the same order, as the time (t,) taken for a significant seasonal change in the environment. Note that in any theory involving continuity, the changes are asymptotic to complete replacement. Thus ideally we may have three classes of cases: 1. t, << t,, competitive exclusion at equilibrium complete before the environment changes significantly. 2. t, 2 t, , no equilibrium achieved. 3. t, >> t,, competitive exclusion occurring in a changing environment to the full range of which individual competitors would have to be adapted to live alone. The first case applies to laboratory animals in controlled conditions, and conceivably to fast breeding bacteria under fairly constant conditions in nature. The second case applies to most organisms with a generation time approximately measured in days or weeks, and so may be expected to occur in the plankton and in the case of populations of multivoltine insects. The third case applies to animals with a life span of several years, such as birds and mammals. Very slow and very fast breeders thus are likely to compete under conditions in which an approach to equilibrium is possible; organisms of inter- mediate rates of reproduction may not do so. This point of view was made clear in an earlier paper (Hutchinson, 1953), but the distribution of that paper was somewhat limited and it seems desirable to emphasize the matter again briefly. It is probably no accident that the great proponents of the type of theory involved in competitive exclusion have been laboratory workers on the one hand (for example, Gause, 1934, 1935; Crombie, 1947; and by implication Nicholson, 1933, 1957) and vertebrate field zoologists (for example, Grin- nell, 1904; Lack, 1954) on the other. The major critics of this type of ap- PARADOX OF THE PLANKTON 139 proach, notably Andrewartha and Birch (1954), have largely worked with insects in the field, often under conditions considerably disturbed by human activity.

DISTRIBUTION OF SPECIES AND INDIVIDUALS MacArthur (1957, 1960) has shown that by making certain reasonable as- sumptions as to the nature of niche diversification in homogeneously diversified' biotopes of large extent, the distribution of species at equilibrium follows a law such that the rth rarest species in a population of Ss species and N, individuals may be expected to be

This distribution, which is conveniently designated as type I, holds remark- ably well for birds in homogeneously diverse biotopes (MacArthur, 1957, 1960), for molluscs of the genus Conus (Kohn, 1959, 1960) and for at least one mammal population (J. Armstrong, personal communication). It does not hold for bird faunas in heterogeneously diverse biotopes, nor for diatoms settling on slides (Patrick in MacArthur, 1960) nor for the arthropods of soil (Hairston, 1957). Using Foged's (19.54) data for the occurrence of planktonic diatoms in Braendegird S# on the Danish island of Funen, it is also apparent (figure 1) that the type I distribution does not hold for such assemblages of diatom populations under quite natural conditions either. MacArthur (1957, 1960) has deduced two other types of distribution (type I1 and type 111) corresponding to different kinds of biological hypotheses. These distributions, unlike type I, do not imply competitive exclusion. So far in nature only type I distributions and a kind of empirical distribution which I shall designate type IV are known. The type IV distribution given by diatoms on slides, in the plankton and in the littoral of Braendegdrd S6, as well as by soil arthropods, differs from the type I in having its commonest species commoner and all other species rarer. It could be explained as due to heterogeneous diversity, for if the biotope consisted of patches in each one of which the ratio of species to individuals differed, then the sum of the assemblages gives such a curve. This is essentially the same as Hairston's (1959) idea of a more structured community in the case of soil arthropods than in that of birds. It could probably arise if the environment changed in favoring temporarily a particular species at the expense of other species before equilibrium is achieved. This is, in fact, a sort of temporal analogue to

'A biotope is said to be homogeneously diverse relative to a group of organisms if the elements of the environmental mosaic relevant to the organism are small compared to the mean range of the organisms. A heterogeneously diverse bzotope is divided into elements at least some of which are large compared to the ranges of the organisms. An area of woodland is homogeneously diverse relative to most birds, a large tract of stands of woodland in open country is heterogeneously diverse (Hutchinson, 1957, 1959). 140 THE AMERICAN NATURALIST

FIGURE 1. Abundance of indbidual species plotted against rank order for the planktonic diatoms of Braendegard Sd, for the four seasons, from Foged's data, showing type IV distributions. The unmarked line gives the type I distribution for a like number of species-and individuals. The unit of population for each species is the ratio of total number of individuals (N,) to total number of species (S,). heterogeneous diversity. Existence of the type IV distribution does not nec- essarily imply non-equilibrium, but if we assume niches are separated out of the niche-hyperspace with any boundary as probable as any other, we may conclude that either non-equilibrium in time or unexpected diversity in space are likely to underlie this type of distribution.

APPLICATION TO THE PLANKTON Before proceeding to inquire how far plankton associations are e~ther never in equilibrium in time or approach heterogeneous diversity in space in a rather subtle way, it is desirable to inquire how far ordinary homogeneous niche diversification may be involved. The presence of a light gradient in all epigean waters by day does imply a certain diversification, but in the epilimnia of lakes the chances of any organism remaining per- manently in a particular narrow range of intensities is small in turbulent water. By day the stability of the epilimnion may well never be zero, but since what has to be explained is the presence of many species of competitors in a small volume of water, the role of small vertical variations is probably insignificant. A few organisms may be favored by peculiar chemical conditions at the surface film, but again this hardly seems an adequate ex- PARADOX OF THE PLANKTON 141 planation. The Langmuir spirals in the wind drift might also separate motile from non-motile forms or organisms of different densities to some extent but again the effect is likely to be small and transitory. It is hard to believe that in turbulent open water many physical opportunities for niche diversification exist.

SYMBIOSIS AND COMMENSALISM The mathematical theory of competition permits the treatment of commensal and symbiotic relations by a simple change in sign of one or both of the competition functions. It can be shown (Gause and Witt, 1935) that under some conditions commensal or symbiotic species can occupy the same niche. There is a little evidence that occasionally water in which one alga has been growing may be stimulatory to another species (Lefsvre, Jacob and Nisbet, 1952; see also Hartman, 1960) though it is far more likely to be in- hibitory. Since some phytoplankters require vitamins and others do not, a more generally efficient species, requiring vitamins produced in excess by an otherwise less efficient species not requiring such compounds, can pro- duce a mixed equilibrium population. It is reasonably certain that this type of situation occurs in the phytoplankton. It is interesting to note that many vitamin-requiring algae are small and that the groups characteristically needing them (Euglenophyta, Crytophyceae, Chrysophyceae, and Dinophy- ceae) tend to be motile. The motility would give such organisms an advantage in meeting rare nutrient molecules, inorganic or organic. This type of advantage can be obtained by non-motile forms only by sinking in a turbulent medium (Munk and Riley, 1952) which is much more dangerous than even random swimming.

ROLE OF PREDATION It can be shown theoretically, as Dr. MacArthur and I have developed in conversation, that if one of two competing species is limited by a predator, while the other is either not so limited or is fed on by a different predator, co-existence of the two prey species may in some cases be possible. This should permit some diversification of both prey and predator in a homogeneous habit.

RESULTS OF NON-EQUILIBRI~MCONDITIONS The possibility of synergistic phenomena on the one hand and of specific predation on the other would probably permit the development of a somewhat diversified equilibrium plankton even in an environment that was essentially boundaryless and isotropic. It may, however, be doubted that such phenomena would ever permit assemblages of the order of magnitude of tens of species to co-occur. At least in homogeneous water in the open ocean there would seem to be no other alternative to a non-equilibrium, or as MacArthur (1960) would term it, an opportunistic community. The great difficulty inherent in the opportunistic hypothesis is that since, if many species are present in a really variable environment which is con- 142 THE AMERICAN NATURALIST trolling their competition, chance extinction is likely to be an important aspect of the process.' That this is not an important aspect of the problem, at least in some cases, is shown by the continual presence of certain dominant species of planktonic diatoms as microfossils in sediments laid down under fairly uniform conditions over periods of centuries or millenia. This is, for instance, clear from Patrick's (1943) study of the diatoms of Linsley Pond, in which locality Stephanodiscus astrea, Melosira ambzgua and certain species of Cyclotella must have co-occurred commonly for long periods of time. It is always possible to suppose that the persistent species were continu- ally reintroduced from outside whenever they became extinct locally, but this does not seem a reasonable explanation of the observed regularity.

IS THE PHYTOPLANKTON A VALID CONCEPT? In view of the paradoxical nature of the phytoplankton, perhaps it is justi- fiable to inquire to what extent the concept itself has validity. In the ocean it is reasonably certain that the community is a self-perpetuating one, but in lakes it has long been regarded as largely an evolutionary derivative of the littoral benthos (for example, Wesenberg-Lund, 1908, pp. 323-325) and in recent years much evidence has accumulated to suggest that the deriva- tion in some cases is not an evolutionary process in the ordinary sense of the word, but a process occurring annually, some individuals of a benthic flora moving at times into plankton. The remarkable work of Lund (1954, 1955) on Melosira indicates that the planktonic species of this genus be- come benthic, though probably in a non-reproductive condition, when turbu- lence is inadequate to keep them afloat. Brook (1959) believes that some of the supposed planktonic varieties of littoral-benthic desmids are non-genetic modifications exhibited by populations annually derived from the littoral. If most of the phytoplankton consisted of species with well-defined, if somewhat restricted, benthic littoral niches, from which at times large cultures in the open water were developed but perhaps left no descendants, much of our paradox would disappear. In the sea we should still apparently have to rely on synergism, predation and opportunistn or failure to achieve equilibrium, but in fresh waters we might get still more diversity from transitory invasions of species which in the benthos probably occupy a heterogeneously diverse biotope like the soil fauna studied by Hairston (1959).

'The chance of extinction is always finite even in the absence of competition, but for the kind of population under consideration the arguments adduced, for instance, by Cole (1960) appear to the writer to be unrealistic. In a lake of area 1 km2 or lo6 m2, in a layer of water only one meter deep, any organism present at a concentration of one individual per litre, which would be almost undetectibly rare to the planktologist using ordinary methods, would have a population No of lo9 individuals. If the individuals divided and the two fission products had equal chances of death or reproduction, so that in the expected case the population remained stable, the probability of random extinction (Skellam, 1955) is given by pe = [t/(l + tjNo where t is measured in generations. For large values of No and t we may approximate by t = -No/lnpe, In the lake in question pe would reach a value of 0.01 in 2.2 x 10' generations which for most phytoplankters would be a period of over a million years. Less than half a dozen lakes are as old as this, and all these are vastly larger than the hypothetical lake of area 1 km2. PARADOX OF THE PLANKTON 143

The available data appear to indicate that in a given lake district there is no correlation between the area of a lake and the number of species com- prising its phytoplankton. This is apparent from JZrneFelt's (1956) monu- mental study of the lakes of Finland, and also From Ruttner's (1952) Fifteen Indonesian lakes. In the latter case, the correlation coefficient of the logarithm of the numbers of phytoplankton species on the logarithm of the area (the appropriate quantities to use In such a case), is -0.019, obviously not significantly different from zero. It is obvious that something is happening in such cases that is quite different from the phenomena of species distribution of terrestrial animals on small islands, so illuminatingly discussed by Dr. E. 0. Wilson in another contribution to this symposium. At first sight the apparent independence in- dicated in the limnological data also may appear not to be in accord with the position taken in the present contribution. If, however, we may suppose that the influence of the littoral on the species composition decreases as the area of the lake increases, while the diversity of the littoral flora that might appear in the plankton increases as the length of the littoral, and so its chances of diversification, increases, then we might expect much less effect of area than would initially appear reasonable. The lack of an observed relationship is, therefore, not at all inconsistent with the point of view here developed.

CONCLUSION Apart from providing a few thoughts on what is to me a fascinating, if somewhat specialized subject, my main purpose has been to show how a certain theory, namely, that of competitive exclusion, can be used to ex- amine a situation where its main conclusions seem to be empirically false. Just because the theory is analytically true and in a certain sense tauto- logical, we can trust it in the work of trying to find out what has happened to cause its empirical falsification. It is, of course, possible that some people with greater insight might have seen further into the problem of the plankton without the theory that I have with it, but for the moment I am content that its use has demonstrated possible ways of looking at the problem and, I hope, of presenting that problem to you.

LITERATURE CITED Andrewartha, H. G., and L. C. Birch, 1954, The distribution and abundance of animals. XV. 782 pp. Univ. of Chicago Press, Chicago, Ill. Brook, A. H., 1959, The status of desmids in the plankton and the determin- ation of phytoplankton quotients. J. Ecol. 47: 429-445. Cole, L. C., 1960, Competitive exclusion. Science 132: 348. Crombie, A. C., 1947, Interspecific competition. J. Animal Ecol. 16: 44-73. Feller, W., 1939, Die Grundlagen der Volterraschen Theorie des Kampfes von Dasein in Wahrscheirlichkeitstheoretischer Behandlung. Acta Biotheoret. 5: 11-40. Foged, N., 1954, On the diatom flora of some Funen Lakes. Folia Limnol. Scand. 5: 1-75. 144 THE AMERICAN NATURALIST

Gause, G. F., 1934, The struggle for existence. IX. 163 pp. Williams and Wilkins, Baltimore, Md. 1935, Vkrifications expkrimentales de la thCorie mathematique de la lutte pour la vie. Actual. scient. indust. 277: 1-62. Gause, G. F., and A. A. Witt, 1935, Behavior of mixed populations and the problem of natural selections. Amer. Nat. 69: 596-609. Grinnell, J., 1904, The origin and distribution of the chestnut-backed Chick- adee. Auk 21: 364-382. Hairston, N. G., 1959, Species abundance and community organization. Ecology 40: 404-416. Hardin, G., 1960, The competitive exclusion principle. Science 131: 1292- 1298. Hartman, R. T., 1960, Algae and metabolites of natural waters. In The Pymatuning symposia in ecology: the ecology of algae. Pymatun- ing Laboratory of Field Biology, Univ. Pittsburgh Special Publ. No. 2: 38-55. Hutchinson, G. E., 1941, Ecolpgical aspects of succession in natural populations. Amer. Nat. 75: 406-418. 1953, The concept of pattern in ecology. Proc. Acad. Nat. Sci. Phila. 105: 1-12. 1957, Concluding remarks. Cold Spring Harbor Symp. Quant. Biol. 22: 415-427. 1959, I1 concetto modern0 di nicchia ecologica. Mem. 1st. Ital. Idrobiol. 11: 9-22. Jsrnefelt, H., 1956, Zur Limnologie einiger Gewgsser Finnlands. XVI. Mit besonderer Beriicksichtigung des Planktons. Ann. Zool. Soc. < 1 Vancimo" 17(1): 1-201. Kohn, A. J., 1959, The ecology of Conus in Hawaii. Ecol. Monogr. 29: 47- 90. 1960, Ecological notes on Conus (Molluscs: Gastropoda) in the Trin- comalee region of Ceylon. Ann. Mag. Nat. Hist. 13(2): 309-320. Lack, D., 1954, The natural regulation of animal numbers. VIII. 343 pp. Clarendon Press, Oxford, England. LefGvre, Id., H. Jakob and M. Nisbet, 1952, Auto- et heteroantagonisme chez les algues d'eau dounce. Ann. Stat. Centr. Hydrobiol. Appl. 4: 5-197. Lund, J. W. G., 1954, The seasonal cycle of the plankton diatom, Meloszra ztalzca (Ehr.) Kiitz. subsp. subarctzca 0. Miill. J. Ec01.42: 151-179. 1955, Further observations on the seasonal cycle of Melosira italica (Ehr.) Kiitz. subsp. subarctica 0. Miill. J. Ecol. 43: 90-102. MacArthur, R. H., 1957, On the relative abundance of bird species. Proc. Natl. Acad. Sci. 45: 293-295. 1960, On the relative abundance of species. Amer. Nat. 94: 25-36. Munk, R1. H., and G. A. Riley, 1952, Absorption of nutrients by aquatic plants. J. Mar. Research 11: 215-240. Nicholson, A. J., 1933, The balance of animal populations. J. Animal Ecol. 2(suppl.): 132-178. 1957, The self-adjustment of populations to change. Cold Spring Har- bor Symp. Quant. Biol. 22: 153-173. PARADOX OF THE PLANKTON 145

Patrick, R., 1943, The diatoms of Linsley Pond, Connecticut. Proc. Acad. Nat. Sci. Phila. 95: 53-110. Provasoli, L., and I. J. Pintner, 1960, Artificial media for fresh-water algae: problems and suggestions. In The Pymatuning symposia in ecology: the ecology of algae. Pymatuning Laboratory of Field Biol- ogy, Univ. Pittsburgh Special Publ. No. 2: 84-96. Rand, A. L., 1952, Secondary sexual characters and ecological competition. Fieldiana Zool. 34: 65-70. Ruttner, F., 1952, Plankton studien der Deutschen Limnologischen Sunda- Expedition. Arch. Hydrobiol. Suppl. 21 : 1-274. Skellam, J. G., 1955, The mathematical approach to population dynamics. In The numbers of man and animals, ed. by J. G. Cragg and N. W. Pirie. Pp. 31-46. Oliver and Boyd (for the Institute of Biology), Edinburgh, Scotland. Udvardy, M. F. D., 1959, Notes on the ecological concepts of habitat, biotope and niche. Ecology 40: 725-728. Wesenberg-Lund, C., 1908, Plankton investigations of the Danish lakes. General part: the Baltic freshwater plankton, its origin and varia- tion. 389 pp. Gyldendalske Boghandel, Copenhagen, Denmark.